rules, but it is not able to synchronize metadata to multiple Apache ShardingSphere instances or be aware of each other in a cluster environment. Updating metadata through one instance causes inconsistencies tables within the same logical table generate repetitive primary keys because they are not mu‐ tually aware. Although collisions can be avoided by constraining the initial value and step size of self‐ 3.1 Logic data source name of readwrite-splitting dynamic-strategy: # Readwrite-splitting type auto-aware-data-source-name: # Database discovery logic data source name write-data-source-query-enabled: #

rules, but it is not able to synchronize metadata to multiple Apache ShardingSphere instances or be aware of each other in a cluster environment. Updating metadata through one instance causes inconsistencies tables within the same logical table generate repetitive primary keys because they are not mu‐ tually aware. Although collisions can be avoided by constraining the initial value and step size of self‐ 3.1 Logic data source name of readwrite-splitting dynamic-strategy: # Readwrite-splitting type auto-aware-data-source-name: # Database discovery logic data source name write-data-source-query-enabled: #

READWRITE_SPLITTING dataSources: (+): # 读写分离逻辑数据源名称 dynamic-strategy: # 读写分离类型 auto-aware-data-source-name: # 数据库发现逻辑数据源名称 write-data-source-query-enabled: # 从库全部下线，主库是否承担读流量 loadBalancerName: rules.readwrite-splitting.data-sources..dynamic-strategy.auto-aware-data-source-name= # 数据库发现逻辑数据源 名称 spring.shardingsphere.rules.readwrite-splitting.data-sources. rules.readwrite-splitting.data-sources..dynamic-strategy.auto-aware-data-source-name= # 数据库发现的逻辑数据 源名称 spring.shardingsphere.rules.database-discovery.data-sources.

Prope rties 读 写 分 离 所 需 属 性， 如 静 态：write‐data‐ source‐name、read‐ data‐sourc e‐names， 动 态：auto‐aware‐ data‐source‐name • loadBala ncerName (?) S tring 读库负载均衡算法名称 轮询负载均衡算法 5.1. ShardingSphere-JDBC dataSources: (+): # 读写分离逻辑数据源名称 type: # 读写分离类型，比如：Static，Dynamic props: auto-aware-data-source-name: # 自动发现数据源名称（与数据库发现配合使用） write-data-source-name: # 写库数据源名称 read-data-source-names: shardingsphere.rules.readwrite-splitting.data-sources..props.auto-aware-data-source-name= # 自动发现数据源名称（与数据库发现配 合使用） spring.shardingsphere.rules.readwrite-splitting.data-sources

rules, but it is not able to synchronize metadata to multiple Apache ShardingSphere instances or be aware of each other in a cluster environment. Updating metadata through one instance causes inconsistencies tables within the same logical table generate repetitive primary keys because they are not mu‐ tually aware. Although collisions can be avoided by constraining the initial value and step size of self‐ incrementing RANDOM Readwrite-Splitting READWRITE_SPLITTING, WRITE_STORAGE_UNIT, READ_STORAGE_UNITS , AUTO_AWARE_RESOURCE Encrypt ENCRYPT, COLUMNS, CIPHER, ENCRYPT_ALGORITHM Database Discovery DB_DISCOVERY,

Prope rties 读 写 分 离 所 需 属 性， 如 静 态：write‐data‐ source‐name、read‐ data‐sourc e‐names， 动 态：auto‐aware‐ data‐source‐name • loadBala ncerName (?) S tring 读库负载均衡算法名称 轮询负载均衡算法 5.1. ShardingSphere-JDBC dataSources: (+): # 读写分离逻辑数据源名称 type: # 读写分离类型，比如：Static，Dynamic props: auto-aware-data-source-name: # 自动发现数据源名称（与数据库发现配合使用） write-data-source-name: # 写库数据源名称 read-data-source-names: shardingsphere.rules.readwrite-splitting.data-sources..props.auto-aware-data-source-name= # 自动发现数据源名称（与数据库发现配 合使用） spring.shardingsphere.rules.readwrite-splitting.data-sources

rules, but it is not able to synchronize metadata to multiple Apache ShardingSphere instances or be aware of each other in a cluster environment. Updating metadata through one instance causes inconsistencies tables within the same logical table generate repetitive primary keys because they are not mu‐ tually aware. Although collisions can be avoided by constraining the initial value and step size of self‐ incrementing RANDOM Readwrite-Splitting READWRITE_SPLITTING, WRITE_STORAGE_UNIT, READ_STORAGE_UNITS , AUTO_AWARE_RESOURCE Encrypt ENCRYPT, COLUMNS, CIPHER, ENCRYPT_ALGORITHM Database Discovery DB_DISCOVERY,

ShardingSphere document 读写分离 READWRITE_SPLITTING、WRITE_STORAGE_UNIT、READ_STORAGE_UNITS 、 AUTO_AWARE_RESOURCE 数据加密 ENCRYPT、COLUMNS、CIPHER、ENCRYPT_ALGORITHM 数据库发现 DB_DISCOVERY、STORAGE_UNITS、HEARTBEAT ‘%s.| | 4 2S02 | 20294 | %s data source name \%s‘ not in database ‘%s.“| | 44000 | 20295 | Auto aware data source name is required in database ‘ ‘%s.“| | 42S02 | 20296 | Not found load balance type 左右可以尽量减少多行结果集 的 fetch 次数。proxy-frontend-executor-size 参数默认值为 CPU * 2，可以根据实际测试结果减 少至 CPU * 0.5 左右；如果涉及 NUMA，可以根据实际测试结果设置为单个 CPU 的物理核数。 server.yaml 文件节选： props: proxy-backend-query-fetch-size: 50 # prox

ShardingSphere document 读写分离 READWRITE_SPLITTING、WRITE_STORAGE_UNIT、READ_STORAGE_UNITS 、 AUTO_AWARE_RESOURCE 数据加密 ENCRYPT、COLUMNS、CIPHER、ENCRYPT_ALGORITHM 数据库发现 DB_DISCOVERY、STORAGE_UNITS、HEARTBEAT ‘%s.| | 4 2S02 | 20294 | %s data source name \%s‘ not in database ‘%s.“| | 44000 | 20295 | Auto aware data source name is required in database ‘ ‘%s.“| | 42S02 | 20296 | Not found load balance type 左右可以尽量减少多行结果集 的 fetch 次数。proxy-frontend-executor-size 参数默认值为 CPU * 2，可以根据实际测试结果减 少至 CPU * 0.5 左右；如果涉及 NUMA，可以根据实际测试结果设置为单个 CPU 的物理核数。 server.yaml 文件节选： props: proxy-backend-query-fetch-size: 50 # prox

ShardingSphere document 读写分离 READWRITE_SPLITTING、WRITE_STORAGE_UNIT、READ_STORAGE_UNITS 、 AUTO_AWARE_RESOURCE 数据加密 ENCRYPT、COLUMNS、CIPHER、ENCRYPT_ALGORITHM 数据库发现 DB_DISCOVERY、STORAGE_UNITS、HEARTBEAT 左右可以尽量减少多行结果集 的 fetch 次数。proxy-frontend-executor-size 参数默认值为 CPU * 2，可以根据实际测试结果减 少至 CPU * 0.5 左右；如果涉及 NUMA，可以根据实际测试结果设置为单个 CPU 的物理核数。 global.yaml 文件节选： props: proxy-backend-query-fetch-size: 50 # prox